the construction and character of a machine or manufacture, there seems to be no meritorious reason for their not having the same effect on the novelty of subsequent patents. No injustice can result from such a rule, because, in order to have any effect on such novelty, drawings as well as words must be able to endure the test stated in the next section. § 57. Novelty is not negatived by any prior patent or printed publication, unless the information contained therein, is full enough and precise enough to enable any person skilled in the art to which it relates, to perform the process or make the thing covered by the patent sought to be anticipated.' It follows from this rule that difference is fatal to anticipation, if that difference is greater than any which could occur between two things, independently constructed by two skilful mechanics from the same specification. The phrase "skilful mechanic," as used in this connection, does not include mechanics who are skilful only in methods of servile imitation. It refers only to mechanics who know how to vary form without varying substance, and who, in constructing a machine from a printed description, or from Patent Office drawings, could readily and would freely alter proportions and change details in order to adapt the contrivance to a particular use, or in order to secure greater merit of workmanship for the thing constructed. If a difference greater than what could arise in this way, is found to exist between a patented thing, and the thing described in a prior patent or printed publication, then the latter cannot negative the novelty of the former. It may, however, defeat the patent for want of invention, though failing to defeat it for want of novelty. Whether or not it can do so, is to be ascertained in each case by applying thereto the established rules on the subject of invention. 1 Seymour v. Osborne, 11 Wallace, 516, 1870; Cawood Patent, 94 U. S. 704, 1876; Downton v. Milling Co. 108 U. S. 466, 1882; Eames v. Andrews, 122 U. S. 66, 1886; Roberts v. Dickey, 4 Fisher, 532, 1871; Cahill v. Brown, 3 Bann. & Ard. 580, 1878; Hammerschlag v. Scamoni, 7 Fed. Rep. 584, 1881. § 58. Novelty is not negatived by any prior abandoned application for a patent.' Abandoned applications for patents are not by the statutes made bars to patents to later applicants. They furnish no evidence that any specimen of the things they describe was ever made or used anywhere. Being only pen and ink representations of what may have existed only as mental conceptions of the men who put them upon paper, they do not prove that the things which they depict were ever known in any country. Nor can they be classed among printed publications, for they are usually in writing, and are not published otherwise than by being placed on file in the Patent Office. $ 59. When there is evidence that he who made and abandoned an application for a patent, made also some effort to carry his invention into practical use, then that application is admissible in evidence to aid the court to determine the date and the nature of the invention which was sought to be embodied in a working form. If, however, upon the whole of the evidence, it appears that what the inventor did, outside of his abandoned application, did not amount to enough to negative the novelty of a subsequent patent to a later inventor, then that abandoned application becomes immaterial to that issue.' § 60. Novelty is not negatived by any successful application for a patent, nor by any documents pertaining thereto, other than the letters patent issued in pursuance thereof.' When such an application, or such a document, is offered to prove the existence of something which is not shown by the letters patent themselves, the justice and propriety of this rule is apparent at a glance. The rule necessarily follows from the same considerations as those which reject an abandoned application, when an abandoned application is offered to negative novelty. But when a successful appli 1 Corn-planter Patent, 23 Wallace, 211, 1874; N. W. Extinguisher Co. v. Phila. Extinguisher Co. 1 Bann. & Ard. 177, 1874; Lyman Ventilating & Refrigerator Co. v. Lalor, 1 Bann. & Ard. 403, 1874. 2 Corn-planter Patent, 23 Wallace, 211, 1874. 3 Howes v. McNeal, 5 Bann. & Ard. 77, 1880. cation is offered only to prove the date of the invention claimed in the resulting patent, a case will occur which will doubtless occasion the establishment of an exception to the rule. § 61. Novelty is not negatived by any prior unpublished drawings, no matter how completely they may exhibit the patented invention,' nor by any prior model, no matter how fully it may coincide with the thing covered by the patent." The reason of this rule is not stated with fullness in either of the cases which support it, but that reason is deducible from the statute, and from the nature of drawings and of models. The statute provides, relevant to the newness of patentable machines and manufactures and improvements thereof, that they shall not have been previously known or used by others in this country." Now, it is clear that to use a model or a drawing, is not to use the machine or manufacture which it represents; and it is equally obvious, that to know a drawing or a model, is not the same thing as knowing the article which that drawing or model more or less imperfectly pictures to the eye. It follows that neither of those things negative the sort of newness required by the statute. Nor is the statutory provision on this point lacking in good reasons to support it. Private drawings may be mislaid or hidden, so as to preclude all probability of the public ever deriving any benefit therefrom; and even if they are seen by several or by many, they are apt to be understood by few or by none. Models also are liable to be secluded from view and to suffer change, and thus to fail of propagation. Moreover, if a patent could be defeated. by producing a model or a drawing to correspond there Ellithorp v. Robertson, 4 Blatch. 309, 1859; Draper v. Potomska Mills, 3 Bann. & Ard. 214, 1878; Detroit Lubricator Mfg. Co. v. Renchard, 9 Fed. Rep. 293, 1881; Odell v. Stout, 22 Fed. Rep. 159, 1884; Pennsylvania Diamond Drill Co. v. Simpson, 29 Fed. Rep. 291, 1886. Cahoon v. Ring, 1 Cliff, 593, 1861; Stainthorp v. Humiston, 4 Fisher, 107, 1864; Johnson v.. McCullough, 4 Fisher, 170, 1870; Stillwell & Bierce Mfg. Co. v.. The Cincinnati Gas Light & Coke Co. 1 Bann. & Ard. 610, 1875. Revised Statutes,. Section 4886. with, and by testifying that it was made at some sufficiently remote point of time in the past, a strong temptation would be offered to perjury. Several considerations of public policy and of private right combine, therefore, to justify the rule of this section. § 62. Novelty is not negatived by anything substantially different from that covered by the patent, even though the function of the prior thing was identical with that of the patented article. This rule necessarily follows from the doctrine that a valid patent may be granted for a new means of producing an old result.' In order to defeat a patent for want of novelty, the anticipating thing must be substantially identical with that covered by the patent. If a marked and apparently important change is found in the patented thing, and one which is not assignable to mere omission of parts, or mere substitution of equivalents,' the defence will fail. But a thing which will not defeat a patent for want of novelty, may defeat it for want of invention. It will do so wherever the observed difference corresponds in character with either of those differences between a patented thing and the prior art, which, in the second chapter of this book, were shown to be insufficient to constitute invention. 2 § 63. The rule of the last section will probably govern every case which justly comes within the doctrine that novelty is not negatived by any unsuccessful abandoned experiment. That rule is far more reliable than that doctrine, because the latter is subject to such qualifications and explanations, that its practical utility in deciding cases is but small. A thing may have been abandoned and still negative the novelty of a thing independently invented long after that abandonment. Such will be the result if the earlier 'O'Reilly v. Morse, 15 Howard, 62, 1853. Jennings v. Lowenstine, 31 Fed. Rep. 84, 1887. 3 Hartshorn v. Barrel Co. 119 U. S. 676, 1886. • Waterman v. Thomson, 2 Fisher, 463, 1863; Shoup . Henrici, 2 Bann. & Ard. 249, 1876; N. W. Extinguisher Co. v. Phila. Extinguisher Co. 1 Bann. & Ard. 177, 1874; McNish v. Everson, 5 Bann. & Ard. 484, 1880. thing was identical with the later, and was used long enough to show that it would work.' If a given experimental device was unsuccessful in the hands of its contriver, that fact must have been due either to one or more faults of principle, or to one or more faults of construction, or to one or more faults of each of these kinds. If partly or wholly due to any fault of principle, that very fact shows that the unsuccessful device was substantially different from subsequent successful patented things. For that reason alone, it would have failed to negative the novelty of those things even if it had not been unsuccessful. If, on the other hand, a prior device was unsuccessful merely because its construction was faulty, it is far from certain that it will not be held to negative the novelty of all subsequent devices identical with it in plan, mode of operation, and function. The Stevens car-brake can conveniently be used to illustrate this point, because the reader can look upon a woodcut of that brake in Webster's Dictionary, and because its construction is particularly suitable to the question in hand. Simple and excellent as that contrivance is, it was not conceived till after several more complicated and less useful brakes were produced and patented. The concep tion came to the inventor in one moment of time, though after a long study of the problem. He knew then, as well as he ever did afterward, that the conceived combination would certainly succeed. Now suppose some other mechanic in the United States had conceived the same apparatus a year earlier, and had promptly constructed all the parts of the first specimen thereof, and had promptly fitted all those parts together and to a railroad car; and suppose that the chain F, being somewhat light in proportion to the strain upon it, had parted asunder the first time the apparatus was attempted to be operated, so that no result was produced; and suppose that, because of 'Gayler v. Wilder, 10 Howard, 477, 1850; Sayles v. Railway Co. 4 Fisher 588, 1871; Stephenson v. |